issues
search
ROCm
/
triton
Development repository for the Triton language and compiler
MIT License
80
stars
22
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
enable atomic MemoryOrderings
#557
scxiao
closed
2 months ago
0
support bypassing data layout conversion for atomic operator
#556
xiaohuguo2023
opened
2 months ago
0
Cherry pick support for tensor pointer usages in scf::IfOp
#555
zhanglx13
closed
3 months ago
0
[tool] Fix occ
#554
zhanglx13
closed
3 months ago
0
[Upstream Backend] [PyTorch UT]: `error: failed to legalize operation 'triton_gpu.local_load' that was explicitly marked illegal`
#553
jataylo
closed
2 months ago
1
[Upstream Backend] [PyTorch UT] `RuntimeError: Triton Error [HIP]: Code: 1, Messsage: invalid argument`
#552
jataylo
closed
2 months ago
5
[FA] Alibi Forward Pass
#551
micmelesse
closed
3 months ago
0
[MFMA] Implement MFMA 4x64 v3
#550
binarman
opened
3 months ago
0
Aot change merge
#549
groenenboomj
opened
3 months ago
0
[Upstream Backend] [PyTorch UT] `error: failed to legalize operation 'tt.mulhiui' that was explicitly marked illegal`
#548
jataylo
closed
2 months ago
7
Add llvm flag
#547
zhanglx13
opened
3 months ago
3
[DotSlicing] Fix AMDReorderInstructionPass
#546
oplavsic
closed
2 months ago
1
fix: replace if/else statement with tl.where
#545
Sara-KS
opened
3 months ago
0
[release/pytorch_2.2] Update required arguments for build scripts
#544
jithunnair-amd
closed
3 months ago
0
Enable fast-exp
#543
zhanglx13
closed
3 months ago
1
[release/pytorch 2.0] Update scripts for wheel build
#542
jithunnair-amd
closed
3 months ago
0
Solve official flash-attention.py fails: typeError:'function' object is not subscriptable
#541
zhangxiao-stack
opened
3 months ago
0
[release/pytorch_2.1] Update required arguments for build scripts
#540
jithunnair-amd
closed
3 months ago
2
[MFMA] MFMA 4x64 64x4 version 2
#539
binarman
opened
3 months ago
0
[MFMA][FRONTEND] Add more options for forced mfma layout sizes
#538
binarman
opened
3 months ago
0
Bias segfault workaround
#537
micmelesse
closed
3 months ago
5
enable register usage and spill for AMD backend
#536
xiaohuguo2023
closed
3 months ago
0
enable register usage and spill for AMD backend
#535
xiaohuguo2023
closed
3 months ago
0
Integrate cudagraph to autotuning
#534
scxiao
closed
3 months ago
6
[AMD] Refactor SharedToDotOperandMFMA
#533
binarman
closed
3 months ago
0
[DotSlicing] Support slicing multiple operands of a load
#532
htyu
closed
3 months ago
1
[DotSlicing] Do not change the frequency of sliced ops.
#531
htyu
opened
3 months ago
0
Add ability to specify custom configs at the cmd line
#530
vgokhale
closed
3 months ago
2
Change critical args to constexprs
#529
vgokhale
closed
4 months ago
0
Added a reference implementation with fake quantization.
#528
wenchenvincent
opened
4 months ago
0
Fixed issues with testing and an issue with amax_o
#527
wenchenvincent
closed
4 months ago
0
[ReductionOp][MFMA] fix reduction of mfma64x4 layout
#526
binarman
closed
3 months ago
0
Vgokhale/causal
#525
vgokhale
closed
4 months ago
0
Fixed an issue when checking second dot
#524
zhanglx13
closed
4 months ago
0
Create main.yml
#523
okakarpa
closed
4 months ago
0
Update amd-offline-tests.yml
#522
okakarpa
closed
4 months ago
0
Create main.yml
#521
okakarpa
closed
4 months ago
0
Added an option to use truncation instead of rounding for fp32 to
#520
wenchenvincent
closed
4 months ago
0
Causal masking with dissimilar Q/KV sequence lengths
#519
vgokhale
closed
4 months ago
0
Merge changes according to AOTriton's testing results.
#518
xinyazhang
opened
4 months ago
0
tl.load: Support passing tl.constexpr(SOME_TUPLE) to boundary_check
#517
xinyazhang
opened
4 months ago
0
[Issue]: Unified memory tensors aren't seen as accessible to Triton
#516
joerowell
closed
1 month ago
5
[Issue]: `nvidia-smi not found`
#515
joerowell
opened
4 months ago
2
triton flash atten module generates wrong results
#514
zhangxiao-stack
closed
1 month ago
20
added bf16 support for perf-kernels
#513
jtang10
closed
4 months ago
2
Unify hasConvertToMMATransisitiveUse
#512
htyu
closed
4 months ago
11
Add a script to run tests in `pytest test_core.py` separately
#511
zhanglx13
closed
4 months ago
0
HIP: Add RUNPATH to compiled so files.
#510
xinyazhang
closed
3 months ago
2
Add causal for nonvarlen tests
#509
vgokhale
closed
4 months ago
0
Head padding for non power of 2 head sizes
#508
groenenboomj
closed
4 months ago
0
Previous
Next